An Improved Method for Vietnam License Plate Location, Segmentation and Recognition

Abstract

The Automatic License Plate Recognition (ALPR) is very important in the Intelligent Transportation System (ITS).In this paper we proposed an improved ALPR algorithm for Vietnam license plates (LP), which consists of three main modules: license plate location (LPL), character segmentation, character recognition. In the location work, we have improved algorithm based on edge detection, image subtraction, mathematic morphology to locate LP region, which considered removing noise. In the segmentation work, we have improved algorithm to get the segments in the LP by the peak-to-valley method in order to segment in digit images getting the two bounds of the each digit according to the statistical parameter. In the recognition work, we have used a Multi Layer Perceptron (MLP) neural network and back-propagation (BP) algorithm to recognize characters & numbers of the Vietnam LP, we used two networks for characters & numbers training with noises, in which the computing time and accuracy is improved.Our approach is more effective than some of the existing methods earlier and satisfied for Vietnam LP. We have been implemented on 600 images taken from actual scenes, different background such as light conditions (night and day), angles, illumination, size and type, colors, reflected light, dynamic conditions. The efficiency of the proposed approach is improved and average rate of accuracy of the one-row LP is 96.93%, two-row LP is 95.82%, higher than most of previous works.

Classification and Comparison of License Plates Localization Algorithms

“…The authors of [11,12] propose a candidate's location by extracting the contours of the binary image using the Canny filter. They added to the opening a dilation operation to connect close objects.…”

Segmentation and Recognition of Korean Vehicle License Plate Characters Based on the Global Threshold Method and the Cross-Correlation Matching Algorithm

“…In this paper, we introduce a new algorithm that combines different image processing algorithms for LPCS, such as image average filtering, visibility restoration, vertical edge-emphasizing, thresholding, morphological operations, and connected component analysis. LPCR is the final and major step after LPCS has been completed in the LPR system Recently, various kinds of optical character recognition (OCR) algorithms have been used for LPCR, such as the template matching technique [12], which is common and very easy to implement; neural networks [13] and support vector machines [14], which are strong and fast classifiers for real-time classification and have significant accuracy; and other least squares-support vector machine (LS-SVM) [15] methods have also been presented for the LPCR process. In this paper, we combined the statistical correlation matching method with the concept of template matching for LPCR.…”